This topic contains 4 replies, has 3 voices, and was last updated by Rene 3 years, 1 month ago.
You must be logged in to reply to this topic.
Welcome Guest View Cart (0 items)Login
Home › Forums › UM7 Product Support › Drift Yaw Angle
This topic contains 4 replies, has 3 voices, and was last updated by Rene 3 years, 1 month ago.
Hello, I’m using the UM7 to stabilize a Quadcopter (own design).
Basically, I set the UM7 to work with Euler Angles and using Tx Rx I send and acquire date from it. The problem I have is the drifting of the yaw angle every 5s approx. I already set the magnetometers and accelerometers (calibrated) using the function that send to the UM7 the value of 0xB0 and 0xAF respectively. However, the problem persist.
If someone can give me a clue how to fix this problem, I would appreciate it.
Thank you.
Hi Alejandro,
It’s most likely that the magnetometer measurements are being distorted by the permanent magnets in the quadrotor motors. On a small rotorcraft like a quadcopter, yaw angle stability is a tough nut to crack.
Thank you Caleb for answer me so fast,
I just fixed the problem. I calibrate the Magnetometers (0xB0), accelerometers (0xAF) , reset the kalman filter (0xAD) and calibrate the gyro (0xAC) before to start the main function (loop). Additionally, I set a function that takes the first values (angles, rates) and then set them like offset (respectively) when the measurement is taken directly from the sensor (without the Euler/Rate factor; 1/91 , 1/16). Then, I reassure that is calibrated.
I have been working since then and is really stable.
I share how I solved in case someone else have the same problem.
Hi Alejandro,
Could you please share your solution with me Im also suffering same kind of problem.. it will be highly appreciated if you can help..
Hi Alejandro,
It seems I do have the same problem as you described.
I would appreciate it if you could share your solution with me.
Best regards,
Rene
You must be logged in to reply to this topic.
Copyright (c) CHRobotics LLC. All rights reserved.